Hewitt reaction revisited

Article abstract of DOI:10.1016/S0040-4039(02)02647-3

A range of alkyl phenylphosphonites are prepared from the reaction of phenylphosphinic acid with the corresponding alkyl chloroformates. A mechanism for this reaction is proposed.

Full text of DOI:10.1016/S0040-4039(02)02647-3

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 781–783
Hewitt reaction revisited
Kamyar Afarinkia* and Hiu-wan Yu
Department of Chemistry, King’s College, Strand, London WC2R 2LS, UK
Received 18 September 2002; revised 15 November 2002; accepted 22 November 2002
Dedicated to Professor Charles W. Rees FRS on the occasion of his 75th birthday
Abstract—A range of alkyl phenylphosphonites are prepared from the reaction of phenylphosphinic acid with the corresponding
alkyl chloroformates. A mechanism for this reaction is proposed. © 2003 Elsevier Science Ltd. All rights reserved.
In 1979 Hewitt published a procedure for the transfor-
mation of phenylphosphinic acid to its monoethyl ester,
compound 1 (Scheme 1).1 The procedure involves the
addition of pyridine to a stirred solution of phenylphos-
phinic acid and ethyl chloroformate in chloroform. This
is followed by an exothermic reaction and rapid effer-
vescence (CO2 gas) after which the reaction is complete.
After a simple workup procedure, the product is
obtained essentially pure and in quantitative yield.
Although this is an exceptionally efficient and clean
method for the preparation of ethyl phenylphosphonite
1, and has been employed on a number of occasions,2–6
neither the mechanism nor the generality of the reac-
tion have ever been investigated. Since efficient and
practically simple methods for the mono-esterification
of phosphinic acids are rare,7 we became interested in
exploiting Hewitt’s reaction for the general synthesis of
mono esters of phenylphosphinic acid. Here, we report
the application of this method to the general prepara-
tion of alkyl phenylphosphonites, as well as the obser-
vations which enable us to propose a mechanism for
this reaction.
To test the generality of the reaction, we carried out
reactions of phenylphosphinic acid with allyl chlorofor-
mate and benzyl chloroformate under Hewitt’s condi-
tions, using chloroform as a solvent. Unexpectedly, the
reactions afforded not only the desired allyl and benzyl
phenylphosphinates, 2 and 3, respectively, but also a
significant quantity of ethyl phenylphosphinate, 1
(Scheme 2). We presumed that the origins of the ethyl
group in 1 must be in the small amount of ethanol
which is added to commercially available chloroform.
Indeed, when these two reactions were carried out in
dichloromethane, which lacks any ethanol as an addi-
tive, we obtained compounds 2 and 3 in excellent yields
Scheme 1.
Scheme 2.
* Corresponding author. E-mail: kamyar.afarinkia@kcl.ac.uk
0040-4039/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)02647-3

782
K. Afarinkia, H. Yu / Tetrahedron Letters 44 (2003) 781–783
Scheme 3.
Scheme 4. Reagents and conditions: (i) toluene, PhP(O)OH, pyridine, reflux; (ii) PhPCl2 7, pyridine, reflux, 1 h, then water; (iii)
MeCO3H in MeCO2H (36%), rt, 2 days.
most electrophiles would react at the more nucleophilic
phosphorus atom, except for those electrophiles (such
as a trimethylsilyl group) that form a considerably
stronger bond with oxygen. An esterification mecha-
nism involving formation of a bond between the oxygen
atom and a carbon electrophile would be unusual, but
explains why the same product is obtained from the
reaction with ethyl chloroformate and diethyl carbon-
ate. At any rate, one possible mechanism could be a
direct bimolecular nucleophilic displacement on the R
group of the chloroformate RO.CO.Cl (Scheme 3). An
alternative mechanism would be through formation of
a ‘mixed anhydride’ species 4 (Scheme 3), formed from
a chloroformate or a carbonate species. Two different
modes for the subsequent steps can be envisaged. Either
an attack of the phosphorus on the oxygen atom (route
A) or attack of an oxygen atom on a carbon atom
with no trace of ethyl phenylphosphinate 1 present.
However, any ethanol present in the reaction mixture
would most likely react with ethyl chloroformate in the
presence of pyridine to afford diethyl carbonate. This
suggests that the alkylating agent in the Hewitt reaction
may be either ethyl chloroformate, EtO.CO.Cl, or
diethyl carbonate, RO.CO.OR generated in situ during
the course of the reaction. Indeed, phenylphosphinic
acid reacts with diethyl carbonate in the presence of
pyridine to afford 1, but at a slightly reduced rate.
Presumably, in the reactions of phenylphosphinic acid
with allyl chloroformate and benzyl chloroformate in
chloroform, some allyl ethyl carbonate and benzyl ethyl
carbonate are formed that act as alkylating agent trans-
ferring either the ethyl or the allyl/benzyl groups.
Proposing a mechanism for Hewitt’s reaction is chal-
lenging. In solution, phenylphosphinic acid exhibits a
tautomerism between a dominant tetracoordinated,
pentavalent phosphinic acid form, and a trivalent, tri-
coordinated phosphinous acid form, which is an
ambident nucleophile. As an ambident nucleophile,
phenylphosphinic acid may react with acylating and
alkylating reagents through both the oxygen and the
phosphorus atoms. However, it would be expected that
Scheme 5. Reagents and conditions: (i) pyridine, CH2Cl2, rt
(caution: effervescence!) then reflux 15 min.

K. Afarinkia, H. Yu / Tetrahedron Letters 44 (2003) 781–783
783
Table 1.
R
Yield (%)
lp (ppm)
lp (PH) (ppm)
JPH (Hz)
1
2
3
4
5
6
7
8
Me
Et
i-Pr
Allyl
Benzyl
Cholestyl
Decyl
73
99
88
80
91
93
85
85
28.3
25.9
23.4
26.3
26.2
23.4
26.1
27.7
7.57
7.57
7.63
7.64
7.65
7.65
7.59
7.66
566
562
559
565
566
560
562
569
2-(Benzyloxy)ethyl
In summary, we have demonstrated that monoesters of
phenylphosphinic acids can be easily and efficiently
prepared from the reaction between phenylphosphinic
acid and a chloroformate in a simple, practical manner.
We are currently investigating further the mechanism of
this interesting and useful reaction and will report our
results in due course.
(mode B). The key difference between the two proposed
routes is on which CꢀO bond is broken. Therefore, we
designed an experiment that would help to resolve the
issue.
The two diastereomeric menthyl phenylphosphinates 5
and 6 were obtained from the reaction of corresponding
(+)-neomenthol and (−)-menthol and dichlorophenyl-
phosphine 7. This reaction is known to proceed
through a O-phosphinylation (a PꢀO bond formation)
and therefore the asymmetry at the carbon atom C-1 is
preserved. Although the phosphorus atom is asymmet-
ric, due to its tautomerism it was not possible to isolate
the diastereomers in a kinetically stable form in either
case (Scheme 4). Therefore, we decided to oxidise the
phosphorus atom to make it a non-asymmetric atom.
The phosphorus atom in the product is no longer
asymmetric and therefore the products of oxidation,
compounds 8 and 9, are diastereomerically pure. We
then carried out the reactions between (+)-neomenthol
chloroformate and (−)-menthyl choroformate, 10 and
11, and phenylphosphinic acid in toluene in the pres-
ence of pyridine. As before, we oxidised the phosphorus
atom.
Acknowledgements
We thank King’s College London Association and Gar-
lick Foundation for a studentship (H.-W.Y.).
References
1. Hewitt, D. G. Aust. J. Chem. 1979, 32, 463.
2. Rima, G.; Satge, J.; Pradel, C.; Grognet, J. M.; Istin, M.;
Sentenacroumanou, H.; Lion, C. Synth. React. Inorg. Met.
Org. Chem. 1992, 22, 277.
3. Rohovec, J.; Vojtisek, P.; Lukes, I. Phosphorus Sulfur and
Silicon and The Related Elements 1999, 148, 79.
4. Sasaki, M. J. Pestic. Sci. 1995, 20, 193.
5. Lei, H. Y.; Stoakes, M. S.; Schwabacher, A. W. Synthesis
1992, 1255.
Analysis of the products from the reactions showed
that the phenylphosphinate esters obtained from reac-
tions between (−)-menthol and dichlorophenylphos-
phine, and between (−)-menthyl choroformate and
phenylphosphinic acid to be identical. However, they
were different from the two identical phenylphosphi-
nate esters obtained from reactions between (+)-neo-
menthol and dichlorophenylphosphine, and between
(+)-neomenthyl choroformate and phenylphosphinic
acid. In other words, the chirality of the ester at C-1 is
not affected during the course of the reaction. This
observation supports our proposed mechanism in which
a PꢀO bond is formed rather than a CꢀO bond.
6. Li, X.; Scott, G. K.; Baxter, A. D.; Taylor, R. J.; Vyle, J.
S.; Cosstick, R. J. Chem. Soc., Perkin Trans. 1 1994, 2123.
7. Dumond, Y. R.; Baker, R. L.; Montchamp, J.-L. Org.
Lett. 2000, 2, 3341.
8. All compounds were fully characterised.
9. Typical experimental procedure. Benzyl phenylphosphinate:
pyridine (9.68 ml, 0.12 mol) was carefully added to a
vigorously stirred solution of benzyl chloroformate (17.13
ml, 0.12 mol) and phenylphosphinic acid (16.8 g, 0.12 mol)
in DCM (250 ml) at room temperature. Once effervescence
had stopped, the solution was refluxed for 15 min, then
allowed to cool to room temperature. The solution was
poured into 0.1 M hydrochloric acid (90 ml) and the
organic layer was separated. After washing with water
(150 ml) and drying over Na2SO4, the solvent was
removed in vacuo to give benzyl phenylphosphinate as a
colourless oil (25.31 g, 91%). lH (500 MHz, CDCl3) 5.12
As has already been alluded to,
a variety of
phenylphosphinate esters can be prepared from the
corresponding chloroformates. To demonstrate further
the generality of the reaction, we also carried out the
reaction of a number of other chloroformates with
phenylphosphinic acid (Scheme 5, Table 1).8,9 The chlo-
roformates derived from a range of primary and sec-
ondary alcohols undergo the reaction and afford the
corresponding phenylphosphinate esters in excellent
yield.
(2H, 2×dd, J1 9, J2 12, OCH6 2Ph), 7.28–7.40 (5H, m,
Aromatic H), 7.49–7.54 (2H, m, aromatic H), 7.58–7.63
(1H, m, aromatic H), 7.65 (1H, d, JP 566, PH), 7.71–7.83
(2H, m, aromatic H); lP (145.67 MHz, CDCl3) 26.22; lc
(125 MHz, CDCl3) 67.64 (OC6 H2Ph), 127.36–133.62 (aro-
matic C); m/z (EI) 232 (90), 167 (917), 141 (9), 126 (87),
107 (55), 91 (100), 79 (89), 65 (22), 51 (12).